New Feature: Introduce __experimentalCreateInterpolateElement

packages/element/CHANGELOG.md

@@ -1,3 +1,7 @@
+## Master
+
+- Added `__experimentalCreateInterpolateElement` function (see [17376](https://github.com/WordPress/gutenberg/pull/17376))
+
 ## 2.8.0 (2019-09-16)
 
 ### New Features

packages/element/src/create-interpolate-element.js

@@ -0,0 +1,325 @@
+/**
+ * External dependencies
+ */
+import { createElement, cloneElement, Fragment, isValidElement } from 'react';
+
+let indoc,
+	offset,
+	output,
+	stack;
+
+/**
+ * Matches tags in the localized string
+ *
+ * This is used for extracting the tag pattern groups for parsing the localized
+ * string and along with the map converting it to a react element.
+ *
+ * There are four references extracted using this tokenizer:
+ *
+ * match: Full match of the tag (i.e. <strong>, </strong>, <br/>)
+ * isClosing: The closing slash, it it exists.
+ * name: The name portion of the tag (strong, br) (if )
+ * isSelfClosed: The slash on a self closing tag, if it exists.
+ *
+ * @type {RegExp}
+ */
+const tokenizer = /<(\/)?(\w+)\s*(\/)?>/g;
+
+/**
+ * Tracks recursive-descent parse state.
+ *
+ * This is a Stack frame holding parent elements until all children have been
+ * parsed.
+ *
+ * @private
+ * @param {WPElement} element          A parent element which may still have
+ *                                     nested children not yet parsed.
+ * @param {number}    tokenStart       Offset at which parent element first
+ *                                     appears.
+ * @param {number}    tokenLength      Length of string marking start of parent
+ *                                     element.
+ * @param {number}    prevOffset       Running offset at which parsing should
+ *                                     continue.
+ * @param {number}    leadingTextStart Offset at which last closing element
+ *                                     finished, used for finding text between
+ *                                     elements
+ *
+ * @return {Frame} The stack frame tracking parse progress.
+ */
+function Frame(
+	element,
+	tokenStart,
+	tokenLength,
+	prevOffset,
+	leadingTextStart,
+) {
+	return {
+		element,
+		tokenStart,
+		tokenLength,
+		prevOffset,
+		leadingTextStart,
+		children: [],
+	};
+}
+
+/**
+ * This function creates an interpolated element from a passed in string with
+ * specific tags matching how the string should be converted to an element via
+ * the conversion map value.
+ *
+ * @example
+ * For example, for the given string:
+ *
+ * "This is a <span>string</span> with <a>a link</a> and a self-closing
+ * <CustomComponentB/> tag"
+ *
+ * You would have something like this as the conversionMap value:
+ *
+ * ```js
+ * {
+ *     span: <span />,
+ *     a: <a href={ 'https://github.com' } />,
+ *     CustomComponentB: <CustomComponent />,
+ * }
+ * ```
+ *
+ * @param {string}  interpolatedString  The interpolation string to be parsed.
+ * @param {Object}  conversionMap       The map used to convert the string to
+ *                                      a react element.
+ * @throws {TypeError}
+ * @return {WPElement}  A wp element.
+ */
+const createInterpolateElement = ( interpolatedString, conversionMap ) => {
+	indoc = interpolatedString;
+	offset = 0;
+	output = [];
+	stack = [];
+	tokenizer.lastIndex = 0;
+
+	if ( ! isValidConversionMap( conversionMap ) ) {
+		throw new TypeError(
+			'The conversionMap provided is not valid. It must be an object with values that are WPElements'
+		);
+	}
+
+	do {
+		// twiddle our thumbs
+	} while ( proceed( conversionMap ) );
+	return createElement( Fragment, null, ...output );
+};
+
+/**
+ * Validate conversion map.
+ *
+ * A map is considered valid if it's an object and every value in the object
+ * is a WPElement
+ *
+ * @private
+ *
+ * @param {Object} conversionMap  The map being validated.
+ *
+ * @return {boolean}  True means the map is valid.
+ */
+const isValidConversionMap = ( conversionMap ) => {
+	const isObject = typeof conversionMap === 'object';
+	const values = isObject && Object.values( conversionMap );
+	return isObject &&
+		values.length &&
+		values.every( ( element ) => isValidElement( element ) );
+};
+
+/**
+ * This is the iterator over the matches in the string.
+ *
+ * @private
+ *
+ * @param {Object} conversionMap The conversion map for the string.
+ *
+ * @return {boolean} true for continuing to iterate, false for finished.
+ */
+function proceed( conversionMap ) {
+	const next = nextToken();
+	const [ tokenType, name, startOffset, tokenLength ] = next;
+	const stackDepth = stack.length;
+	const leadingTextStart = startOffset > offset ? offset : null;
+	if ( ! conversionMap[ name ] ) {
+		addText();
+		return false;
+	}
+	switch ( tokenType ) {
+		case 'no-more-tokens':
+			if ( stackDepth !== 0 ) {
+				const { leadingTextStart: stackLeadingText, tokenStart } = stack.pop();
+				output.push( indoc.substr( stackLeadingText, tokenStart ) );
+			}
+			addText();
+			return false;
+
+		case 'self-closed':
+			if ( 0 === stackDepth ) {
+				if ( null !== leadingTextStart ) {
+					output.push(
+						indoc.substr( leadingTextStart, startOffset - leadingTextStart )
+					);
+				}
+				output.push( conversionMap[ name ] );
+				offset = startOffset + tokenLength;
+				return true;
+			}
+
+			// otherwise we found an inner element
+			addChild(
+				new Frame( conversionMap[ name ], startOffset, tokenLength )
+			);
+			offset = startOffset + tokenLength;
+			return true;
+
+		case 'opener':
+			stack.push(
+				new Frame(
+					conversionMap[ name ],
+					startOffset,
+					tokenLength,
+					startOffset + tokenLength,
+					leadingTextStart
+				)
+			);
+			offset = startOffset + tokenLength;
+			return true;
+
+		case 'closer':
+			// if we're not nesting then this is easy - close the block
+			if ( 1 === stackDepth ) {
+				closeOuterElement( startOffset );
+				offset = startOffset + tokenLength;
+				return true;
+			}
+
+			// otherwise we're nested and we have to close out the current
+			// block and add it as a innerBlock to the parent
+			const stackTop = stack.pop();
+			const text = indoc.substr(
+				stackTop.prevOffset,
+				startOffset - stackTop.prevOffset
+			);
+			stackTop.children.push( text );
+			stackTop.prevOffset = startOffset + tokenLength;
+			const frame = new Frame(
+				stackTop.element,
+				stackTop.tokenStart,
+				stackTop.tokenLength,
+				startOffset + tokenLength,
+			);
+			frame.children = stackTop.children;
+			addChild( frame );
+			offset = startOffset + tokenLength;
+			return true;
+
+		default:
+			addText();
+			return false;
+	}
+}
+
+/**
+ * Grabs the next token match in the string and returns it's details.
+ *
+ * @private
+ *
+ * @return {Array}  An array of details for the token matched.
+ */
+function nextToken() {
+	const matches = tokenizer.exec( indoc );
+	// we have no more tokens
+	if ( null === matches ) {
+		return [ 'no-more-tokens' ];
+	}
+	const startedAt = matches.index;
+	const [ match, isClosing, name, isSelfClosed ] = matches;
+	const length = match.length;
+	if ( isSelfClosed ) {
+		return [ 'self-closed', name, startedAt, length ];
+	}
+	if ( isClosing ) {
+		return [ 'closer', name, startedAt, length ];
+	}
+	return [ 'opener', name, startedAt, length ];
+}
+
+/**
+ * Pushes text extracted from the indoc string to the output stack given the
+ * current rawLength value and offset (if rawLength is provided ) or the
+ * indoc.length and offset.
+ *
+ * @private
+ */
+function addText() {
+	const length = indoc.length - offset;
+	if ( 0 === length ) {
+		return;
+	}
+	output.push( indoc.substr( offset, length ) );
+}
+
+/**
+ * Pushes a child element to the associated parent element's children for the
+ * parent currently active in the stack.
+ *
+ * @private
+ *
+ * @param {Frame}    frame       The Frame containing the child element and it's
+ *                               token information.
+ */
+function addChild( frame ) {
+	const { element, tokenStart, tokenLength, prevOffset, children } = frame;
+	const parent = stack[ stack.length - 1 ];
+	const text = indoc.substr( parent.prevOffset, tokenStart - parent.prevOffset );
+
+	if ( text ) {
+		parent.children.push( text );
+	}
+
+	parent.children.push(
+		cloneElement( element, null, ...children )
+	);
+	parent.prevOffset = prevOffset ? prevOffset : tokenStart + tokenLength;
+}
+
+/**
+ * This is called for closing tags. It creates the element currently active in
+ * the stack.
+ *
+ * @private
+ *
+ * @param {number} endOffset Offset at which the closing tag for the element
+ *                           begins in the string. If this is greater than the
+ *                           prevOffset attached to the element, then this
+ *                           helps capture any remaining nested text nodes in
+ *                           the element.
+ */
+function closeOuterElement( endOffset ) {
+	const { element, leadingTextStart, prevOffset, tokenStart, children } = stack.pop();
+
+	const text = endOffset ?
+		indoc.substr( prevOffset, endOffset - prevOffset ) :
+		indoc.substr( prevOffset );
+
+	if ( text ) {
+		children.push( text );
+	}
+
+	if ( null !== leadingTextStart ) {
+		output.push( indoc.substr( leadingTextStart, tokenStart - leadingTextStart ) );
+	}
+
+	output.push(
+		cloneElement(
+			element,
+			null,
+			...children
+		)
+	);
+}
+
+export default createInterpolateElement;

packages/element/src/index.js

@@ -1,3 +1,4 @@
+export { default as __experimentalCreateInterpolateElement } from './create-interpolate-element';
 export * from './react';
 export * from './react-platform';
 export * from './utils';